Input data available for download

Posted on March 23, 2012

We are pleased to inform you that a first set of bias-corrected climate input data is now available for use in your ISI-MIP impact simulations.

The data is kindly hosted by the Climate Service Center (CSC) on a server in Hamburg, Germany. To access it, please go to, and the "Input data". There you will find some important information related to the data download; please be sure to read all of it carefully before proceeding.

Further down on the same page, you will find a link to the server where the data is located. On the server, you can browse the available climate models, scenarios, and variables, and download the desired files. For download, you will need to log in with the following user information:
username: XXX
password: XXX
(Please do not confuse this with the user account for our website that was previously distributed for the download of dummy input data. That account has been deactivated now that the proper input data is available; all relevant information on our website is freely accessible.)

Please note that currently only a limited set of data is available: historical data and two RCPs for the HadGEM2-ES model; as well as historical data and one RCP for the IPSL-CM5A-LR model. The provision of a more comprehensive data set has been delayed mainly by the fact that data is still incomplete for the majority of climate models in the CMIP5 archives. We are aware that the limited data set we are providing now falls far short of your (and our) expectations, and we apologize for this.

We are working hard to obtain, and then pre-process, additional climate model data. We will be adding the new data to the server as it becomes available, and we will keep you informed about any changes. In the meantime, we encourage you to start downloading the data that is available, and to prepare and run your simulations according to the Simulation Protocol as soon as you have the required input data. We also remind you to provide us a first output data set for test purposes by May 1st.

Please also note the remarks regarding missing data, as well as known issues regarding data quality, as documented on our website ( under "Input data". We will be updating this information as new issues come to our attention, and will try to fill gaps and fix bugs wherever possible.

Finally, please do not hesitate to contact us if you have any questions regarding the input data; in particular, if any problems come to your attention during download and/or use of the data, please do let us know.

Dear ISI-MIP participants,

It has come to our attention that a mistake has occurred in the preparation of the precipiation data for the WATCH historical dataset.  The original version of the data published by us on the DRKZ server contained the variable pr, which mistakenly only included rainfall.  This was a mistake since we define pr as the total precipitation (rainfall + snowfall). 

New files have been prepared (for the observational period as well for the spin-up), where pr contains total precipitation, and prsn contains the snowfall component only, and are now available on the DRKZ server.

We apologise sincerely for this inconvenience, especially for those who have already been using the WATCH data to run their simulations.  Should this mean that it is no longer realistic for you to submit your data by the 30 November deadline, please get in touch and let us know. 

Best wishes,
The ISI-MIP team.